The project has two Specific Aims: SA1. a) Quantify and characterize the parametric variation of cerebral response to delay, probability, and reward magnitude during decision making in AD vs Controls. b) Quantify disruptions of cerebral connectivity as a correlate of increased impulsivity in AD. SA2. Determine the extent to which relapse can be predicted by phenotypic variation in the neural substrates of behavioral impulsivity. METHODS: Eighty recently abstinent AD subjects will be recruited from local residential treatment programs. Forty control subjects will be recruited from the general public and non alcohol dependent friends of the patients. Subjects will be followed for 3 months after their initial sobriety date. They will be evaluated monthly with computerized and questionnaire measures of impulsivity. Patients will keep a drinking diary to allow capture of all elements of relapse during the study period. They will undergo a functional and anatomical MRI evaluation at the initial visit to quantify the neural response to cardinal elements of impulsive choice and to assess the extent of functional and anatomical changes in cerebral connectivity. A statistical model will relate information from brain imaging to the likelihood of relapse. RELEVANCE: The neural profiles defined in this study will improve understanding of the cardinal elements of discounting behavior as a predictor of relapse in AD. Quantitative profiles of mechanisms of choice in AD individuals will provide valuable information for the tailoring of individual treatment regimens. This will directly impact future treatment approaches not only in alcoholism but also for substance abuse and addictive behaviors in general. This knowledge will allow funding for substance abuse treatment programs to be used more effectively.